Name: ai Code Review Strategist
Price: 10 USD
Availability: InStock
Author: Agensi

See it in action

You say

Review this PR for the payment processing module. It was generated by an AI agent and I need to make sure it doesn't have any hidden hallucinations or silent failures.

Your agent does

PR: Payment Module — Review Summary

Blast Radius

Files changed: 2
Downstream consumers: 14
Critical path: Yes
Review depth: Deep

AI Failure Mode Findings

[SILENT_ERROR] payment.py:45

Issue: catch Exception block logs and returns None, swallowing database timeouts.
Severity: Critical

Verdict: REJECTED

The error handler fails to propagate transaction failures, risking data inconsistency. Re-implement with explicit exception types.

What you get

Identify hallucinated API calls in AI-generated pull requests.Detect silent error swallowing in critical system paths.Scale engineering lead review capacity for high-velocity teams.Verify that AI-generated tests cover more than just the happy path.

About this skill

Teaches AI coding agents to perform structured, high-signal code reviews specifically for AI-generated code — catching the failure modes unique to LLM output (confident hallucinations, silent error swallowing, hallucinated API contracts, missing edge cases, over-optimized style with wrong logic). Implements a tiered review framework: blast radius analysis, AI-failure-mode heuristics, fitness function verification, and structured output. For senior engineers drowning in AI-generated PRs that "look correct" but break in production.